Probabilistic analysis of georisk from capping of contaminated sediment in Gunneklev Fjord, Norway

The seabed of the Gunneklev Fjord in south-eastern Norway is covered with exceptionally soft contaminated sediment containing mercury and dioxins. The sediments have an undrained shear strength less than 1 kPa and a variable thickness of up to 2.5 m. To reduce the potential for leaching of contaminants from these sediments, Hydro Energy AS developed a remediation plan involving capping of the sediment. Significant uncertainties are associated with the sediment's undrained shear strength and the cap thickness. An unfortunate combination of low shear strength and thick cap could cause slope failure (e.g. translational sliding of the sediment and/or bearing capacity failure (e.g. punching through). Failure, particularly slope failure, can cause spreading of the contaminants in the fjord causing serious consequences. This paper presents an assessment of the probability of slope failure associated with the contaminated sediment before and after the cap placement. Probability of bearing capacity failure is considered to have more local effect and is also discussed briefly. The study used the Monte Carlo method with random undrained shear strength and sediment thickness. The simulations show high slope failure probability when the seabed inclination was >= 1:50 combined with an average undrained shear strength <= 0.4 kPa. Based on the probabilistic analyses, a pilot field testing campaign was carried out (after this study) with 20-cm thickness cap in the "gentle" sloping area and 5-cm thickness in the "steep" areas to avoid causing a failure. The analyses were used to support evidence-based decision-making on the cap design and implementation for further field testing.